Branch: refs/heads/bolo/amdgpu_call
Home: https://github.com/dyninst/dyninst
Commit: b12b128af291b20d9b90ce2a32c8b4a01fbac21f
https://github.com/dyninst/dyninst/commit/b12b128af291b20d9b90ce2a32c8b4a01fbac21f
Author: Bolo <bolo@xxxxxxxxxxx>
Date: 2023-08-01 (Tue, 01 Aug 2023)
Changed paths:
A instructionAPI/src/amdgpu_callinsn_table.h
Log Message:
-----------
Dyninst continues parsing amdgpu binaries after a procedure call.
Previously, dyninst would not "fall through" instruction parsing
at a procedure call instruction. This has been changed to allow
a continuation edge for further parsing of a binary.
AMD GPU Call instructions are broken out into a seperate include
file that can be updated from the automation, to avoid further
hand-edits as more AMDGPU architectures are supported.
Commit: 005ddd02eb1d4af7baf12e2246573e079497ad47
https://github.com/dyninst/dyninst/commit/005ddd02eb1d4af7baf12e2246573e079497ad47
Author: Bolo <bolo@xxxxxxxxxxx>
Date: 2023-08-01 (Tue, 01 Aug 2023)
Changed paths:
M instructionAPI/src/InstructionCategories.C
M instructionAPI/src/amdgpu_branchinsn_table.h
A instructionAPI/src/amdgpu_callinsn_table.h
Log Message:
-----------
Dyninst continues parsing amdgpu binaries after a procedure call.
Previously, dyninst would not "fall through" instruction parsing
at a procedure call instruction. This has been changed to allow
a continuation edge for further parsing of a binary.
AMD GPU Call instructions are broken out into a seperate include
file that can be updated from the automation, to avoid further
hand-edits as more AMDGPU architectures are supported.
Commit: 2733adbfa588a896fe5ba85e8deedd2873e7778c
https://github.com/dyninst/dyninst/commit/2733adbfa588a896fe5ba85e8deedd2873e7778c
Author: Bolo <bolo@xxxxxxxxxxx>
Date: 2023-08-01 (Tue, 01 Aug 2023)
Changed paths:
Log Message:
-----------
Merge branch 'bolo/amdgpu_call' of github.com:dyninst/dyninst into bolo/amdgpu_call
Compare: https://github.com/dyninst/dyninst/compare/b12b128af291%5E...2733adbfa588
|